DE19801773A1 - Service quality criterion mapping method for communication network - Google Patents

Abstract

The method involves mapping transparently a first low priority data stream from a connection oriented communication network in a connectionless communication network. A first high priority data stream is mapped transparently in the connectionless communication network. A second low priority data stream from the connectionless communication network is mapped in the connection orientated communication network. An address of the connectionless network is converted into an address of the connection orientated network. A second high priority data stream from the connectionless network is mapped in the connection orientated network. If the second high priority data stream is a signalling packet a conversion is carried out to convert a signalling information from the connectionless network into a signalling information provided by the connection oriented network. If the second high priority data stream is an information packet a transparent transmission to the connection oriented communication network is carried out.

Description

The invention relates to a method for imaging a a connection-oriented communication network Quality of Service feature that has a high priority data stream of a low-priority data stream into one connectionless communication network and vice versa.

A local area network, hereinafter referred to as LAN, is known from [1]. In a LAN, data is transmitted through the Transport of data packets (data of limited length) transfer. The data packets contain the origin address and destination address and are spontaneous, d. H. without one previous connection establishment, sent (connectionless Communication).

A quality of service characteristic denotes one option from several possible options for a given communication network are characteristic. For example, one is below Operating mode or an additional service of the Understand communication network. A quality of service feature in the ATM network is the identifier for the transmission of a high-priority data stream, i.e. a data stream that follows certain criteria (see below) secure transmission (see below) requires of a low priority data stream, which accordingly does not require secure transmission.

Under a transparent transmission or one transparent data stream is a transmission or a Understand data stream, its content without an analysis is transmitted. The implementation of the Content in tax information and data. This corresponds to the  the expert is familiar with the OSI layer model according to the ISO definition (ISO = International Standardization Organization). A stratified view is possible communication technology beyond the OSI Layer model expedient and usual.

From [2] is an ATM network (ATM = Asynchronous Transfer Mode) known. The ATM network works connection-oriented, i. H. data transmission consists of establishing a connection, an information transfer and a disconnection. Further details can be found in [2] and numerous others Sources and the corresponding standards are taken.

It is known to use a LAN, e.g. B. Ethernet or Token Ring to one Connect ATM network. There are predefinable changes to make the definition of the LAN, it is in this Connection of a LAN emulation, still as LANE designated, spoken. For more detailed information to the corresponding LANE definition of the "ATM Forum" referenced (see [3]).

The LANE described in [3] has the disadvantage that none Use of the quality of service features defined in the ATM network (English: Quality of Service, QoS) is possible. Because such Quality of service characteristics in the connected LANs, e.g. B. the Ethernet, also not available, does not exist Possibility of different treatment of Data streams, i.e. a distinction between lower priorities and high priority data streams, as is common in the ATM network is.

Now it is possible to use the existing LANE or that to expand conventional LAN in that within the LAN makes such a distinction between low-priority and high-priority Data streams can be hit. Such an extension can the patent application PCT / DE97 / () 0404 from 04.03.1997 are removed and will be repeated again below  summarized. Such an extended LAN is as follows referred to as LANE2.

So far, however, the quality of service characteristics go at every transition lost between LANE2 and ATM network.

The object of the invention is one in one connection-oriented communication network available Quality of service characteristic, namely the distinction between one high priority and a low priority data stream, in one map connectionless communication network and vice versa.

This object is achieved in accordance with the features of patent claim 1 solved.

The invention forms a connection-oriented Quality of service available in a communications network connectionless communication network. Example of a connection-oriented communication network is above mentioned ATM network and for a connectionless Communication network with the LANE also described above a corresponding extension.

A first low-priority data stream or a first one high-priority data stream becomes your connection-oriented Communication network transparent in the connectionless Transfer communication network. A second low priority Data flow is generated from the connectionless communication network transferred into the connection-oriented communication network, wherein an address translation of an address of the connectionless Communication network in an address of the connection-oriented communication network becomes. A second high priority data stream is created from the connectionless communication network in the connection-oriented communication network mapped by if the second high priority data stream is one Signaling package is an implementation of a  Signaling information of the signaling package from the connectionless communication network in one in the connection-oriented communication network available Signaling information is performed, or if there is the second high-priority data stream Information package deals, a transparent transmission in the connection-oriented communication network becomes.

A first development of the invention describes the implementation of service quality parameters available in the connectionless communication network into the connection-oriented communication network in the following steps:

• - it becomes a from the available bandwidth determine appropriate bandwidth for a new connection;
• - from the address of the connectionless communication network an associated address of the connection-oriented communication network determined;
• - if a suitable bandwidth is available and the Address of an associated participant in the connection-oriented communication network is known, the new connection is established.

Another training is that connection-oriented communication network is an ATM network and the address of the connection-oriented Communication network is an ATM address.

An additional training is that connectionless communication network is a LAN, the Address of the connectionless communication network a LAN Adress is. This LAN corresponds to the one mentioned above the quality of service feature high priority / low priority data stream extended variant LANE2.  

A further development consists in the design of the Address translation (from the LAN to the ATM network). For one specifiable LAN address, an ATM address is found by either the ATM address on a LANE server or via specially built connections to all neighboring nodes is available.

Further developments of the invention also result from the dependent claims.

Exemplary embodiments of the Invention presented in more detail.

Show it

Fig. 1 is a sketch that illustrates the functionality of mapping a quality of service feature from the ATM network to the LAN (and back) in logical blocks.

Fig. 2 is a sketch, the three arrangements that are part of a communication relationship, shows in a layer model (according to ISO),

Fig. 3 is a sketch that shows various protocol formats of the Ethernet standard and an extension according to an RMAC concept.

DESCRIPTION OF THE KNOWN EXTENSION OF THE LAN STANDARD FOR DIFFERENTIATION OF A HIGH PRIOR AND A LOW PRIOR DATA FLOW (SEE PATENT APPLICATION PCT / DE97 / 00404)

In the following, as indicated above, the content of the Patent application PCT / DE97 / 00404 from 04.03.1997 summarized.

In FIG. 2, a first array A1, a second array A2, and a third arrangement A3 are shown, wherein the first arrangement AI with the second array A2 via a channel K1 and the second array A2 to the third arrangement A3 via a channel K2 are connected are.

Both the first arrangement A1 and the third arrangement A3 represent end devices that are based on the Ethernet standard work. The second arrangement shows an Ethernet Switching node (Ethernet switch). Via channels K1 and K2 are data packets (information packets and / or Signaling packets) sent.

According to the Ethernet standard, standardized layers according to ISO (International Standardization Organization) are shown in FIG. 2 for each arrangement. The channels K1 and K2 are coupled via a physical layer PHY (physical layer). An Ethernet layer MAC (Media Access Control) is located above the physical layer PHY, which guarantees the functionality of the standardized Ethernet protocol.

An extension of the Ethernet standard consists in inserting a layer RMAC (= Realtime Media Access Control) between the Ethernet layer MAC and the higher layers HL (network and transport protocols, e.g. TCP / IP (Transport Control Protocol / Internet Protocol) or UDP / IP (User Datagram Protocol / Internet Protocol))
The second arrangement A2 in FIG. 2 is the Ethernet switching node and has no higher layers HL, since these are not necessary for switching functionality.

The RMAC layer performs a. the service, a data packet assign a priority class, the data packet is marked accordingly and is finally transmitted. A recipient, e.g. B. the third arrangement A3 determines the Priority class of the data packet and transmits if necessary the data packet further or provides in the data packet information contained in the higher layers HL transparent to disposal.  

There are preferably four priority classes:

• 1. A first priority class PK0 is suitable for connectionless data transmission. The data packet is without Establishing a connection and no guarantee of successful Transfer (no confirmation) transferred. The first Priority class PK0 represents the lowest priority that can be assigned to a data packet.
• 2. A second priority class PK1 corresponds to one connection-oriented transmission, in which a controlled delay in the transmission of the Data packets guaranteed in a statistical mean The following service features are implemented:
• - it becomes a bandwidth necessary for the connection, which is determined during connection establishment, ensured;
• - an average transmission delay of the data packet, even when the Communication network, is no worse than one Delay of a data packet that the Priority class PK0 has been assigned;
• - The greatest possible delay of the data packet (with the Priority class PK1) is less than the delay a data packet of priority class PK0;
• - a loss rate in the transmission of data packets the priority class PK1 is irrelevant as long as conditions agreed when establishing the connection for the communication link is held.
• 3. A third priority class PK2 concerns one connection-oriented transmission. The following are for that Features provided:
• - most data packets are transmitted completely;
• - for most data packets, one is replaced by a predeterminable threshold specified maximum Delay time not exceeded.
• 4. Finally, a fourth priority class corresponds to PK3, which is also for connection-oriented transmission  is provided, the highest priority that one Data packet can be assigned. It should definitely secure transmission can be ensured, whereby the following features are provided:
• - The negotiated during the connection establishment Bandwidth is provided for transmission;
• - There are predeterminable maximum delay times guaranteed;
• - There are no losses of data packets due to Overflows in buffer stores.

Thus, as long as the communication network cannot collapses when a data packet is assigned to Priority class PK3 from a secure and the fastest possible transmission can be assumed.

Present in terms of connecting the LAN the ATN network only differentiated two priority classes: data with high priority as a high priority data stream and data with lower priority than low priority data stream.

According to the assignment described above Priority class for a data packet also becomes a Labeling of the data packet made, the Ethernet standard format with a corresponding Extension for the RMAC concept.

In Fig. 3 the data packet in different formats and with different content (and functionalities) is shown.

Fig. 3a shows an in Ethernet standard defined data packet with the following fields:

• - Destination address field DA (Destination Address);
• - Sender address field SA (Source Address);
• - TYPE field:
  The TYPE field contains an integer, preferably two octets long. Depending on the value, the meaning of the TYPE field is interpreted differently:
  • - TYPE <1500: interpretation as a length field (data packet corresponds to an LLC format (Logical Link Control));
  • - TYPE ≧ 1500: interpretation as type specification (code for the network protocol);
• - An INFO field contains further service features or data or signaling information that normally be transmitted transparently to the higher layers HL;
• - A field FCS contains a checksum for error correction.

The LLC format is shown in Fig. 3c. In addition to the fields shown above, the following fields are also available:

• a field LEN corresponds to the length field (the field TYPE from FIG. 3a contains a value which is less than 1500, thus the field TYPE migrates to the field LEN);
• - provides a field DSAP (Destination Service Access Point) a service access point to the recipient;
• - A field SSAP (Source Service Access Point) contains one Service access point of the transmitter;
• - A CONTROL field contains one or two octets and is used for special command and response functions, where a sequence number is the order of the messages ensures.

The extension of the Ethernet standard with regard to the RMAC concept is shown by way of example in FIG. 3b. For this purpose, a priority mark PM is entered in the field TYPE, which uniquely identifies the priority class PKi (i = {0, 1, 2, 3}).

A data packet with the lowest priority class PK0 is preferably not provided with a priority mark. It is transmitted unchanged in accordance with the Ethernet standard.  

A field F: [D (flow identifier) is also shown in FIG. 3b. This flow identifier uniquely identifies the sending arrangement. Therefore, the receiver, preferably the higher layers HL there, of this data packet provided with the field FID can establish a clear reference to the sending higher layers HL.

The data packet can be an information packet or a Signaling package.

Fig. 3c and Fig. 3d show such signaling packets, wherein in Fig. 3d, the expansion in the sense of the RMAC concept to the Ethernet standard in FIG. 3c is shown. The priority mark PM is contained in the field DSAP. The priority mark PM is also DSAP or SSAP information with additional priority information.

The meanings of the designations of the individual fields in FIG. 3d can be seen from the above explanations.

EMBODIMENT (S) OF THE INVENTION

The data stream present in the Ethernet or Token Ring (LAN) should also be treated in accordance with the specifiable quality of service characteristics when transitioning to the ATM network. In Fig. 1 is a general architecture is shown ATM network for mapping quality of service features of an expanded in the above-described sense (RMAC) conventional LANs in a. Here, 1 is a sketch in Fig. Specified, combines the logical functions in blocks without a division of hardware components and / or software components can be seen from this illustration. Such a division is left to the respective specific application and can vary accordingly.

The individual components (blocks) in FIG. 1 are explained below.

A block DF ("Data Forwarders") distinguishes between one high priority and a low priority data stream. Act it a low-priority data stream, then a Further processing in a block LEC ("LAN Emulation Client") initiated (see description below). Is it on the other hand, a high-priority data stream is a Further processing in a block TETAS ("Transparent Ethernet to ATM slot ") if it is a Otherwise, the information package deals with one Signaling package, a further treatment in one Block RM ("Resource Manager") initiated.

In the block RM resources, e.g. B. the bandwidth, managed. The aim is to allocate resources fairly guarantee. For this purpose, the block in the RM Signaling package contained quality of service parameters from the LAN, preferably adapted from the RMAC protocol, in such a way that these quality of service parameters are implemented in the ATM network can be and are thus available in the ATM network.

Is the establishment of a connection (from the LAN to the ATM network) not possible, a corresponding message will be sent to the System that requested the connection is transmitted (Error handling).

If the resource (channel, bandwidth) is available, the connection with the corresponding quality of service parameters built up. To do this, a LAN address is converted into an ATM address converted to find the target on the ATM side. Is the ATM address is not available locally, so it must come from the block ARU ("Address Resolution Unit") can be determined.

In the block LEC all operations for low priority are Packets in accordance with the LANE specification carried out. This will make it compatible with the standardized LAN extension, which in turn, like  mentioned at the beginning, about no adaptation to the Quality of service features of the ATM standard, ensured. The LEC block manages its own resources, these self-managed resources the block RM be communicated.

In the TETAS block, transparent information packets from transferred to the LAN in the ATM network and back.

Finally, as mentioned, the address implementation takes place in the Block ARU instead. There is a MAC address (Media Access Control) converted into an ATM address. The address translation takes place in two stages. First, a server Accessed component of the LAN, the LAN Emulation Server (LES) to a standardized LANE address resolution perform. No address information from the LES is returned in a second step by ATM Side, preferably there in the exchanges that corresponding ATM address requested. For this, the desired MAC address directly in the B-HLI field ("Broadband High Layer Information "), there in octet 6 to 11 of the Setup message coded (compare [4]). It is only in the Switch that knows the MAC address that Connection accepted. All other exchanges reject the connection.

A block SIG ("signaling") and a block ATMPS ("ATM Protocol Stack ") identify the ATM side, the block ATMPS a standardized ATM layer including the AAL-5 layer (see [5]). The block SIG contains the Signaling protocols Q.2931 and SSCOP (compare [6]).  

The following publications have been cited in this document:
[1] Gerd Sigmund: Basics of mediation technology, 2nd edition, v. Decker, Heidelberg 1992, ISBN 3-7685-4892-9, pp. 177-185.
[2] Gerd Sigmund: Basics of mediation technology, 2nd edition, v. Decker, Heidelberg 1992, ISBN 3-7685-4892-9, pp. 420-428.
[3] ATM Forum: LAN Emulation over ATM V1.0, ATM Forum LANE- 0021.000, January 1995, especially chapter 7 (pages 86-96) and chapter 8 (pages 96-102).
[4] ATM Forum: ATM User Network Interface, Version 3.1.
[5] ITU-T: I.363 B-ISDN ATM adaptation layer (AAL) specification.
[6] ITU-T: Q.2931 Digital Subscriber Signaling System No.2 - User Network Interface (UNI) layer 3 specification for basic call / connection control.
ITU-T: Q.2110 B-ISDN signaling ATM adapation layer - service specific connection oriented protocol (SSCOP).

Claims (5)

1. A method for mapping a quality of service feature available in a connection-oriented communication network, which characterizes a high-priority data stream or a low-priority data stream, into a connectionless communication network and vice versa,

• a) in which a first low-priority data stream from the connection-oriented communication network is mapped transparently into the connectionless communication network,
• b) in which a first high-priority data stream from the connection-oriented communication network is mapped transparently into the connectionless communication network,
• c) in which a second low-priority data stream is mapped from the connectionless communication network into the connection-oriented communication network, an address conversion of an address of the connectionless communication network into an address of the connection-oriented communication network is carried out,
• d) in which a second high-priority data stream is mapped from the connectionless communication network into the connection-oriented communication network by
  if the second high-priority data stream is a signaling packet, conversion of signaling information from the signaling packet from the connectionless communication network into signaling information available in the connection-oriented communication network is carried out,
  if the second high-priority data stream is an information package, a transparent transmission is carried out in the connection-oriented communication network.

2. The method according to claim 1, in which the quality of service parameters available in the connectionless communication network are converted from the signaling packet into the connection-oriented communication network in the following steps:

• a) a suitable bandwidth for a new connection is determined from an available total bandwidth;
• b) the address of the connectionless communication network is determined by an address conversion, the suitable address of the connection-oriented communication network;
• c) a suitable bandwidth is available and if the address is known in the connection-oriented communication network, the new connection is established;
• d) if no suitable bandwidth is available or if the address is not known in the connection-oriented communication network, a predefinable error message is displayed.

3. The method according to any one of the preceding claims, where the connection-oriented communication network ATM network and the address of the connection-oriented Communication network are an ATM address. 4. The method according to any one of the preceding claims, where the connectionless communication network is a local Area Network (LAN), which is about the quality of service feature, a high priority and a low priority data stream distinguish, is expanded, and the address of the connectionless communication network a LAN Address are.   5. The method according to claim 3 and 4, wherein the address conversion is carried out in the following steps:

• a) for the LAN address, the ATM address is determined by a LAN server, or,
• b) if the ATM address is not available locally, a request is made to an ATM server, the ATM server transmitting the ATM address in response.